An International Publisher for Academic and Scientific Journals
Author Login 
Scholars Journal of Physics, Mathematics and Statistics | Volume-13 | Issue-10
Purity in Module Theory and Its Relatives: A Survey of Foundations, Recent Advances and Open Problems
Ram Ji Pandey, Ashok Kumar Pandey*
Published: Nov. 18, 2026 | 12 7
Pages: 331-349
Downloads
Abstract
Cohn introduced purity for modules as a tensor-theoretic weakening of splitting. Since then, it has developed into a family of related notions, used in classical ring theory as well as in exact, Grothendieck and triangulated categories. In this survey we organise the family around one question: which short exact sequences should be treated as almost split, and what do the corresponding relative injective, projective and flat modules say about the ring? We first set up a common language of proper classes defined by test modules, and place cyclic, relative-divisibility, neat, simple-module and torsion-theoretic purities, with their copure duals, in it. We then report on work from the last four years: purities attached to multiplicative subsets and their uniform versions, strong purity and S-phantom morphisms, pure-direct-injective and pure-direct-projective modules, Mittag-Leffler and definability methods, and purity in exact and triangulated categories. Primeness of submodules and of ideals in near-rings is discussed as well, because several recent papers treat both topics. The papers of the second author on torsion purity, T-purity, divisibility, cyclic purity and σ-purity are compared with this literature. Along the way we prove a few short results that we could not find in print: a relative form of Chase’s theorem on products of flat modules, a torsion-theoretic form of Warfield’s theorem on Prüfer domains, the existence of envelopes for τ-injective modules and covers for τ-divisible modules, and a reduction of relative pure semisimplicity to the absolute case. A count of the one hundred references by theme and period shows where the activity is concentrated. We end with twelve open problems, each with its motivation and a suggested approach.